Connect ChatGPT to ECGrid

There are two ways to use ECGrid capabilities in ChatGPT, depending on your goal.

Approach	What it gives you
Custom GPT (Actions)	Full ECGrid REST API inside a purpose-built GPT
OpenAI Responses API	Programmatic ECGrid MCP tool access for developers building apps

Option 1 — Custom GPT with ECGrid REST API (Recommended)

A Custom GPT is a purpose-built version of ChatGPT that can call the ECGrid REST API directly using OpenAI's Actions feature. You import the ECGrid OpenAPI spec, add your API key, and ChatGPT can look up mailboxes, query interchanges, check partner status, and perform any other ECGrid operation — no MCP required.

Who can use this: ChatGPT Plus, Pro, Team, Business, and Enterprise accounts.

Step 1 — Open the GPT Editor

In chat.openai.com, click your profile picture → My GPTs → Create a GPT → Configure tab.

Give your GPT a name and description, for example:

Name: ECGrid Assistant
Description: Helps manage EDI trading partners, mailboxes, interchanges, and parcels on ECGrid.

Step 2 — Add the ECGrid Action

In the Configure tab, scroll to Actions and click Create new action.

In the Schema field, import or paste the ECGrid OpenAPI spec URL:

https://rest.ecgrid.io/swagger/v2/swagger.json

Click Import from URL. ChatGPT will fetch and parse the spec, listing all available ECGrid REST endpoints.

Step 3 — Configure Authentication

After importing the schema, click Authentication and configure:

Field	Value
Authentication type	`API Key`
API key	Your ECGrid API key
Auth type	`Custom`
Custom header name	`X-APIKey`

:::caution Keep your API key secure Enter your API key in the Authentication dialog — never paste it into the GPT's instructions or the schema. OpenAI stores it encrypted and does not expose it in the GPT configuration UI. :::

Step 4 — Set Instructions (Optional)

In the Instructions box you can give the GPT context about ECGrid, for example:

You are an ECGrid EDI assistant. ECGrid is a B2B EDI network that routes electronic
trading documents between businesses. You can look up mailboxes, trading partners,
interchanges, and parcels. When a user asks about EDI traffic, use the ECGrid API
tools to retrieve live data. Always show network and mailbox IDs when displaying results.

Step 5 — Save and Test

Click Save (top right) → choose Only me or Anyone with the link for access.

Open your new GPT and test with prompts like:

What mailboxes are on network 7?
Show me the interchange status for interchange 123456789
Are there any pending parcels in mailbox 142's inbox?

What the Custom GPT Can Do

The ECGrid REST API has 121 endpoints across 16 categories. Through the Custom GPT Actions your GPT can:

Look up mailboxes, networks, trading partners, and comm records
List inbox/outbox parcels and interchanges
Query interchange status and document counts
Check callback registrations and carbon copy rules
Retrieve user and key/value data

For the full endpoint list, see the ECGrid REST API reference.

Option 2 — OpenAI Responses API (Developers)

If you are building an application using the OpenAI Responses API, you can connect it directly to the ECGrid MCP server by passing X-APIKey as a custom header. Headers are supplied per-request alongside the tool configuration.

Python
JavaScript

from openai import OpenAI

client = OpenAI()  # uses OPENAI_API_KEY from environment

response = client.responses.create(
    model="gpt-4o",
    tools=[
        {
            "type": "mcp",
            "server_label": "ecgrid",
            "server_url": "https://mcp.ecgrid.io/mcp",
            "headers": {
                "X-APIKey": "YOUR_ECGRID_API_KEY"
            },
            "require_approval": "never"
        }
    ],
    input="List the pending parcels in mailbox 142 on network 7"
)

print(response.output_text)

import OpenAI from "openai";

const client = new OpenAI(); // uses OPENAI_API_KEY from environment

const response = await client.responses.create({
  model: "gpt-4o",
  tools: [
    {
      type: "mcp",
      server_label: "ecgrid",
      server_url: "https://mcp.ecgrid.io/mcp",
      headers: {
        "X-APIKey": "YOUR_ECGRID_API_KEY",
      },
      require_approval: "never",
    },
  ],
  input: "List the pending parcels in mailbox 142 on network 7",
});

console.log(response.output_text);

:::caution Headers per request The Responses API does not store headers between requests. Supply X-APIKey in the headers field on every call. :::

The Responses API path gives you programmatic access to all 41 ECGrid MCP tools with the full tool schema, resources, and prompts.

Option 1 — Custom GPT with ECGrid REST API (Recommended)​

Step 1 — Open the GPT Editor​

Step 2 — Add the ECGrid Action​

Step 3 — Configure Authentication​

Step 4 — Set Instructions (Optional)​

Step 5 — Save and Test​

What the Custom GPT Can Do​

Option 2 — OpenAI Responses API (Developers)​

See Also​